New functions: MSE and MSSS

Hi @aho and @erifarov,

I have included two new functions in this branch:

• MSE to compute the Mean Squared Error
• MSSS to compute the Mean Squared Error Skill Score

These functions are very similar to RMSE and RMSSS (same metrics without the squared root), but @nperez and I think it is important to include them as they are used in several studies instead of the RMSE/RMSSS.

I have slightly modified RMSE and RMSSS:

• It is now possible to have memb_dim in RMSE to compute the ensemble mean inside (it was already possible in RMSSS)
• In RMSSS, I have move the lines that created the ref array in case it is provided as a number (e.g. 0 to use the standard climatological forecast). The reason is that it failed when checking that it has the same dimensions as exp because it was still a number instead of an array.
• I have added two more examples to both functions.

Please let me know if anything is unclear or if you want to discuss anything.

Thank you,
Carlos